On the choice of an electromyogram data compression method

Electromyogram (EMG) data compression is of great importance within the framework of telemedicine. For example, there is an increasing demand in medicine to achieve patient healthcare directly from the office of the specialist. The aim of the research presented in this paper is to investigate several kinds of compression methods applied to EMG signals in order to find the method that is most well-suited to EMG data compression. Thus, in this paper, the application of several compression methods to EMG data is studied: predictive linear methods, transform methods and, more specifically, methods based on the wavelet transform. Each method is briefly discussed and experimental results are presented in terms of the signal-to-noise ratio and the compression ratio. The results show that methods based on the wavelet transform outperform the other compression methods